The Impact of Minimum Legal Drinking Age Laws on Alcohol Consumption
Total Page:16
File Type:pdf, Size:1020Kb
The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: Evidence from a regression discontinuity design using exact date of birth Barış K. Yörük† Ceren Ertan Yörük‡ University at Albany, SUNY Northeastern University Abstract This paper uses a regression discontinuity design to estimate the impact of the minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use among young adults. Using data from the National Longitudinal Survey of Youth (1997 Cohort), we find that granting legal access to alcohol at age 21 leads to an increase in several measures of alcohol consumption, including an up to a 10 percentage point increase in the probability of drinking. Furthermore, this effect is robust under several different parametric and non-parametric models. We also find some evidence that the discrete jump in alcohol consumption at age 21 has negative spillover effects on marijuana use but does not affect the smoking habits of young adults. Our results indicate that although the change in alcohol consumption habits of young adults following their 21st birthday is less severe than previously known, policies that are designed to reduce drinking among young adults may have desirable impacts and can create public health benefits. Keywords: alcohol consumption, marijuana use, minimum legal drinking age, smoking JEL classification: I10, I18, I19 † Corresponding author. Department of Economics, University at Albany, SUNY. E-mail: [email protected] ‡ Department of Economics, Northeastern University. E-mail: [email protected] The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: Evidence from a regression discontinuity design using exact date of birth January 24, 2011 Abstract This paper uses a regression discontinuity design to estimate the impact of the minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use among young adults. Using data from the National Longitudinal Survey of Youth (1997 Cohort), we find that granting legal access to alcohol at age 21 leads to an increase in several measures of alcohol consumption, including an up to a 10 percentage point increase in the probability of drinking. Furthermore, this effect is robust under several different parametric and non-parametric models. We also find some evidence that the discrete jump in alcohol consumption at age 21 has negative spillover effects on marijuana use but does not affect the smoking habits of young adults. Our results indicate that although the change in alcohol consumption habits of young adults following their 21st birthday is less severe than previously known, policies that are designed to reduce drinking among young adults may have desirable impacts and can create public health benefits. Keywords: alcohol consumption, marijuana use, minimum legal drinking age, smoking JEL classification: I10, I18, I19 1Introduction While consuming alcohol sensibly is generally associated with better health and longer life, the abuse of alcohol is associated with many undesirable health outcomes. For instance, several studies link heavy alcohol consumption with low blood pressure, increased risk of stroke, liver diseases, and 1 kidney failure.1 Furthermore, the benefits of moderate drinking are found to be mostly small, while the damage caused by heavy drinking appears to be considerable.2 A large body of literature also documents considerable spillover effects of alcohol consumption on labor market outcomes, risky behavior, alcohol related traffic injuries and fatalities, and criminal activity.3 Given these direct and indirect effects of alcohol use, evaluating the effectiveness of the policies regulating alcohol availability and consumption is vital. Many studies have shown that policies that increase the full cost of consuming alcohol such as restricting the days of alcohol sales, toughening drunk driving laws, and raising alcohol taxes significantly decrease alcohol consumption and have positive spillover effects on alcohol consumption related outcomes. One of the most direct forms of regulation on alcohol availability in the United States is imposing a minimum legal drinking age (hereafter, MLDA) of 21. Understanding the effect of the MLDA is particularly important not only because alcohol consumption is related to several undesirable health and economic outcomes, but also lowering the MLDA from 21 is a current policy debate in many states. Proponents of the MLDA of 21 argue that imposing an age limit encourages a majority of young adults under age 21 to consume alcohol in an irresponsible manner and that lowering the drinking age would help young adults to learn how to drink gradually, safely and in moderation. However, opponents of lowering the drinking age argue that states that previously lowered the drinking age to 18, such as Massachusetts, Michigan, and Maine, experienced an increase in alcohol-related traffic accidents among the 18 to 20 age group. Although several studies have investigated the effect of the MLDA laws on alcohol consumption, most of them have made use of the changes in the MLDA that occurred in the 1970s and 1980s at the state level. However, states where a lower MLDA was imposed might be different in unobserved ways than those states where the MLDA of 21 was enforced. If these unobserved differences at the state level are also associated with drinking habits of young adults, than one cannot estimate a consistent effect of the MLDA on alcohol consumption and alcohol consumption related outcomes using the simple variation of the MLDA law at the state level. In order to address this shortcoming, we exploit the discontinuity in drinking habits of young adults at age 21 andusearegressiondiscontinuity 1 See, for example, Hansagi et al. (1995), Mann, Smart, and Govoni (2003), and Niholson and Taylor (1940). 2 Poikolainen (1996) discusses the relationship between alcohol consumption and overall health outcomes in detail. 3 According to the recent National Highway Safety Administration data, in 2008, 37 percent of traffic fatalities in the United States were alcohol related. In recent papers, Carpenter and Dobkin (2008) and Markowitz (2005) document the relationship between alcohol consumption and crime. 2 (hereafter, RD) design to estimate the causal effect of the MLDA on alcohol consumption, smoking, and marijuana use among young adults. Our main identifying assumption is that the observed and unobserved determinants of these outcomes are likely to be distributed smoothly across the age-21 cutoff.4 Hence, the changes in alcohol consumption, smoking, and marijuana use trends at age 21 can solely be attributed to the MLDA law itself. We use a restricted version of the National Longitudinal Survey of Youth, 1997 Cohort (NLSY97) for the empirical analysis, which contains a unique information on the exact birth dates of the respon- dents. In the context of a RD design, this unique information is particularly important since one can clearly identify the treatment and control groups and compare the drinking, smoking, and marijuana use habits of young adults who are just about to turn 21 with those who recently turned 21.5 This paper makes two main contribution to the existing literature. First, using a restricted version of NLSY97, our study provides new estimates of the effect of the MLDA on alcohol consumption behavior of young adults. Similar to the previous studies, our results suggest that granting legal access to alcohol at age 21 leads to an increase in several measures of alcohol consumption. In particular, we find that the MLDA of 21 is associated with up to a 10 percentage point increase in the probability of alcohol consumption and a more than 1.7 day increase in the number of days that young adults consume alcohol per month. We also document that the MLDA increases the probability of binge drinking up to 8 percentage points under different parametric specifications. However, in contrast to previous literature, we find that MLDA does not significantly effect the number of drinks that young adults had on the days they consumed alcohol. Furthermore, the MLDA increases youths’ average alcohol consumption per day by only 0.2 drinks under certain specifications, which suggests that the effect of the MLDA on drinking habits of young adults is less severe than previously known. Second, we investigate the relationship between alcohol consumption and two alcohol consumption related outcomes, namely smoking and marijuana use. The existing literature provides mixed results for the possible relationship between alcohol consumption and these two alcohol related outcomes. 4 This assumption is partially testable. We present the relevant tests in section 5. 5 Suppose that one has information only on the month and year of the birthdate of each respondent and her interview date. Then, treatment and control groups cannot be cleary identified. For instance, a respondent who was born on January 30, 1980 and interviewed on January 1, 2001 will be mistakenly coded as a 21 year old and placed in a treatment group (those who are 21 and older). But, this respondent is actually in the control group since she is 29 days younger than 21 at the time of the interview. Furthermore, by definition, the RD approach estimates the local treatment effect, which calls for a very detailed information around the age-21 cutoff. 3 We find some evidence that the discrete jump in alcohol consumption at age 21 has negative spillover effects on marijuana use. In particular, our results imply that under certain specifications, the proba- bility of marijuana use among young adults tend to increase up to 7 percentage points at age-21 cutoff. However, we find no significant effect of the MLDA on smoking habits of young adults. Furthermore, in general, these results are robust under several parametric or non-parametric specifications. The rest of this paper is organized as follows. The next section provides a summary of the history of the MLDA laws in the United States and discusses the relevant research. Section three presents the data and discusses the relationship between the MLDA and alcohol consumption, marijuana use, and smoking.